We borrowed a lot of code from these repos to build model and engine, thanks for these great works and the open-source community!

• espnet

• Apache-2.0 License
• python/shell utils
• kaldi feat preprocessing
• data pipe line and transformer
• some tts models, like fastspeech2 and GAN-based vocoder

• wenet

• Apache-2.0 License
• U2 model
• Building TLG based Graph
• websocket server & client

• kaldi

• Apache-2.0 License
• shell/perl/python utils.
• feature bins.
• WFST based decoding for LM integration.

• delta

• Apache-2.0 License
• engine arch

• speechbrain

• Apache-2.0 License
• ECAPA-TDNN SV model
• ASR with CTC and pre-trained wav2vec2 models.

• chainer

• MIT License
• Updater, Trainer, and some utils.

• librosa

• ISC License
• Audio feature

• ThreadPool

• zlib License
• ThreadPool

• g2pW

• Apache-2.0 license

*transformers

• Apache-2.0 License
• Wav2vec2.0